Dynamic combination method and system of flexible intelligent manufacturing unit based on modular design

Abstract

This application relates to the field of intelligent manufacturing execution systems (MES) technology, and discloses a method and system for dynamic combination of flexible intelligent manufacturing units based on modular design. The method includes: pre-configuring standard solutions covering common tasks; acquiring production tasks containing task type and target preference data; generating combination methods using a strategy of prioritizing standard solutions and dynamically combining and supplementing them, and comprehensively considering time, cost, quality, and cycle time matching to calculate performance indicators and select the optimal combination; performing rapid static checks through a pre-set conflict graph, followed by dynamic process verification through virtual simulation, conducting two-level screening of candidate combinations to ensure feasibility and safety; finally, issuing instructions to execute physical reassembly after successful simulation verification. This invention enables rapid, efficient, reliable, and intelligent combination of intelligent manufacturing units.

Description

Technical Field

[0001] This application relates to the field of intelligent manufacturing execution system technology, and in particular to a method and system for dynamic combination of flexible intelligent manufacturing units based on modular design. Background Technology

[0002] As market demand shifts towards personalization, diversification, and small-batch production, traditional large-scale, rigid production models are no longer adequate. Flexible Manufacturing Systems (FMS), which utilize modular, reconfigurable manufacturing cells (such as robots, CNC machine tools, AGVs, and inspection equipment) to address changing production tasks, have become a key development direction for intelligent manufacturing.

[0003] In practical applications, the core challenge lies in how to quickly, efficiently, and reliably combine these independent modular devices into a temporary or semi-permanent production line according to different production tasks. Existing technical solutions mainly suffer from the following drawbacks: Dynamic combination using exhaustive or heuristic algorithms for each new task, while flexible, involves enormous computational demands, leading to long decision-making times, slow response times, and difficulty in meeting the demands of fast-paced production; complete reliance on fixed production lines results in a loss of flexibility in responding to task changes; many combination methods prioritize minimizing total processing time as the sole optimization objective, neglecting equally important business and engineering factors such as manufacturing costs, processing quality, and equipment energy consumption; furthermore, they ignore actual bottlenecks and equipment idleness caused by mismatched cycle times in sequential processes, making the theoretically optimal solution inefficient in actual production; while the designed device combination scheme may be theoretically feasible, physical conflicts are possible. Simple static checks cannot detect complex spatial interference, dynamic motion conflicts, and communication timing competition. Direct physical reassembly and debugging without verification can easily lead to equipment collisions, production delays, or even safety accidents, resulting in high trial-and-error costs. Therefore, there is an urgent need for a flexible intelligent manufacturing unit dynamic combination method that can balance response speed and flexibility, make multi-objective and realistic optimization decisions, ensure successful deployment through efficient and reliable verification methods, and possess self-learning and evolutionary capabilities.

[0004] Therefore, the present invention provides a method and system for dynamic combination of flexible intelligent manufacturing units based on modular design. Summary of the Invention

[0005] This application provides a method and system for dynamic combination of flexible intelligent manufacturing units based on modular design, which can improve production efficiency and reduce trial and error costs.

[0006] Firstly, this application provides a method for dynamic combination of flexible intelligent manufacturing units based on modular design, the method comprising: Step S1: Pre-configure multiple standard schemes. Each standard scheme corresponds to a task type. The standard scheme contains fixed equipment combinations, process parameters, processing time, conversion time data, and baseline cycle time values. Step S2: Obtain production task information. Production task information includes the type of task to be processed, the number of tasks of each type, and target preference data. Target preference data includes production time weight, manufacturing cost weight, and processing quality weight. Step S3: Based on the production task information, standard scheme and standardized capability data of each equipment, generate at least one combination method, calculate the performance index of each combination method, select the combination method with the best performance index as the target combination method, and select this combination method as the target combination method if only one combination method is generated. Step S4: Detect the target combination method based on the pre-constructed conflict map. If there is a conflict in the target combination method, select the combination method with the second best performance index as the target combination method until there is no conflict in the selected target combination method. Step S5: Simulate the operation process of the selected target combination method in a virtual environment to verify its feasibility. After the simulation verification is passed, generate control commands according to the target combination method and send control commands to each device to execute the physical layer reorganization.

[0007] Secondly, this application provides a flexible intelligent manufacturing unit dynamic combination system based on modular design, the system comprising: The pre-configuration module is used to pre-configure multiple standard solutions. Each standard solution corresponds to a task type. The standard solution contains fixed equipment combinations, process parameters, processing time, conversion time data, and baseline cycle time values. The task acquisition module is used to acquire production task information, which includes the type of task to be processed, the number of tasks of each type, and target preference data. The target preference data includes production time weight, manufacturing cost weight, and processing quality weight. The combination optimization module is used to generate at least one combination method based on production task information, standard schemes and standardized capability data of each equipment, calculate the performance index of each combination method, select the combination method with the best performance index as the target combination method, and select this combination method as the target combination method when only one combination method is generated. The conflict detection module is used to detect target combination methods based on a pre-built conflict map. If there is a conflict in the target combination method, the combination method with the second best performance index is selected as the target combination method until the selected target combination method has no conflict. The simulation deployment module is used to simulate the operation of the selected target combination in a virtual environment to verify its feasibility. After the simulation verification is successful, control commands are generated according to the target combination and sent to each device to execute the physical layer reorganization.

[0008] Compared with the prior art, the beneficial effects of the present invention are at least as follows: The technical solution provided in this application adopts a two-stage strategy of prioritizing standard solutions and dynamically combining alternatives. For common tasks, pre-optimized standard solutions can be directly called to achieve millisecond-level rapid response. Dynamic combinations with large computational loads are only initiated when facing new tasks, ensuring flexibility while shortening decision time in most cases. By introducing target preference data that includes production time, manufacturing cost, and processing quality weights, the single engineering optimization problem is transformed into a configurable, multi-objective comprehensive decision problem. By introducing a cycle time penalty mechanism in the calculation of performance indicators, the degree of cycle time mismatch among units in the combination is quantified and penalized, effectively avoiding equipment idleness and overall inefficiency caused by the barrel effect. This ensures that the selected combination solution not only has a short theoretical processing time but also has higher equipment utilization and output efficiency in actual continuous production. A two-level verification process of coarse screening with conflict graph and fine screening with virtual simulation is constructed. The conflict graph eliminates combinations with static infeasibility with high efficiency, and the dynamic interaction process is accurately reproduced through virtual simulation to discover potential hidden problems such as motion interference and timing competition, maximizing the success rate of physical reorganization and avoiding expensive physical trial and error costs and safety risks. Attached Figure Description

[0009] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic diagram of an embodiment of the flexible intelligent manufacturing unit dynamic combination method based on modular design in this application. Figure 2 This is a schematic diagram of an embodiment of the flexible intelligent manufacturing unit dynamic combination system based on modular design in this application. Detailed Implementation

[0011] This application provides a method and system for the dynamic combination of flexible intelligent manufacturing units based on modular design. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0012] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the dynamic combination method for flexible intelligent manufacturing units based on modular design in this application includes: Step S1: Pre-configure multiple standard schemes. Each standard scheme corresponds to a task type. The standard scheme contains fixed equipment combinations, process parameters, processing time, conversion time data, and baseline cycle time values. Specifically, to address the issues of long computation times and slow response caused by the need to dynamically reconfigure equipment for each changing production task, multiple standard schemes covering common task types are pre-configured to solidify commonly used combinations, preparing for rapid selection later. Based on the analysis of historical production data and common task types, engineers pre-design several standard schemes. A standard scheme is a predefined template of a set of equipment combinations and their operating modes. A standard scheme can handle one or more task types and internally specifies which equipment is needed, how the equipment coordinates, the processing time for each workpiece, the time required to switch from one piece of equipment to the next, and the production rhythm (baseline cycle time) of the bottleneck equipment in the standard scheme. Equipment combination refers to the list of all physical equipment in the standard scheme, such as lathes, robots, and testing instruments. Process parameters refer to the set values ​​of each piece of equipment, such as processing programs and operating speeds. Processing time refers to the standard time taken to complete a complete process under the standard scheme. Conversion time data refers to the time taken to switch from one piece of equipment to another. For a standard scheme, the baseline cycle time refers to the single-piece processing time of the slowest piece of equipment that limits the overall output speed in the standard scheme.

[0013] Step S2: Obtain production task information. Production task information includes the type of task to be processed, the number of tasks of each type, and target preference data. Target preference data includes production time weight, manufacturing cost weight, and processing quality weight. Specifically, different production tasks have different efficiency requirements. If task preferences are not considered and the same standard is used every time, such as only pursuing the shortest time, cost or quality may be ignored, resulting in a combination that does not meet actual needs. In order to obtain the specific requirements of the current production task and the direction that should be prioritized when making decisions (time, cost or quality), so that subsequent steps can select the optimal combination, for example, if a production task is received that requires processing 50 disc parts and 30 shaft parts, the target preference data in the production task information includes a production time weight of 0.7, a manufacturing cost weight of 0.2, and a processing quality weight of 0.1, by clarifying the target preference of the production task, the subsequent dynamic combination can be adjusted according to the actual needs of the user, improving flexibility and practicality.

[0014] Step S3: Based on the production task information, standard scheme and standardized capability data of each equipment, generate at least one combination method, calculate the performance index of each combination method, select the combination method with the best performance index as the target combination method, and select this combination method as the target combination method if only one combination method is generated. Specifically, production tasks may involve multiple task types. A standard solution usually only covers one or a few similar tasks. Therefore, it may be necessary to chain multiple standard solutions into a combination, or when standard solutions are insufficient, to construct new combinations from the equipment pool. In order to solve the problem of how to quickly select an execution solution that meets the task requirements and is efficient from many possibilities, at least one feasible combination is generated and a unified index is calculated to obtain the optimal target combination. If only one standard solution exists that can cover all the task types that need to be processed, then there is no need to combine multiple standard solutions and make subsequent comparisons. The standard solution that meets the requirements can be directly selected as the target combination. If a set of standard solutions exists that can cover all task types (a set of standard solutions refers to a combination of multiple standard solutions that can cover all task types), then all possible combinations are enumerated, a first indicator is calculated, and a performance indicator is obtained by considering the cycle time penalty. The combination with the best performance indicator is selected as the target combination. If no such combination exists, then dynamic equipment combination is performed: all feasible equipment sequences are enumerated from the candidate equipment according to the task sequence, a second indicator is calculated for each sequence based on the target preference data, and a performance indicator is obtained based on the second indicator combined with the cycle time penalty. The combination with the best performance is selected as the target combination. Through the above steps, a two-stage solution of prioritizing standard solutions and dynamically combining to fill gaps is achieved. This ensures efficient response to common tasks, handles newly emerging tasks, and ensures the selection of the optimal target combination by comparing different combinations through unified performance indicator calculation.

[0015] Step S4: Detect the target combination method based on the pre-constructed conflict map. If there is a conflict in the target combination method, select the combination method with the second best performance index as the target combination method until there is no conflict in the selected target combination method. Specifically, virtual simulation is computationally expensive. Simulating every candidate combination would significantly reduce overall efficiency. Conflict graphs, based on static knowledge, can quickly eliminate a large number of infeasible combinations, acting as a coarse screen. To address the issue that combinations may be optimal in terms of performance metrics but are practically undeployable, conflict graphs are used for rapid pre-screening, eliminating obviously unworkable combinations and avoiding wasting time in subsequent simulations. The conflict graph is pre-built. During detection, all devices in the target combination are extracted to generate device pairs. The conflict graph is then checked for conflict records. If an intolerable conflict is found, the corresponding combination is abandoned, and the next best performing combination is selected as the target combination. Conflict detection is then performed on the newly selected target combination. If conflicts still exist, the combination with the lowest performance metrics is selected from the previously generated combinations in descending order of performance metrics until the selected target combination is free of conflict.

[0016] Step S5: Simulate the operation process of the selected target combination method in a virtual environment to verify its feasibility. After the simulation verification is passed, generate control commands according to the target combination method and send control commands to each device to execute the physical layer reorganization.

[0017] Specifically, conflict maps only record static, known conflicts. However, dynamic interactions between devices, such as instantaneous interference at specific time sequences, may not be reflected in the conflict map. Simulation can simulate behavior on the real timeline. To address the issue that even if a combination method passes the conflict map detection, it may still fail due to dynamic interactions, a final verification is performed through simulation. This ensures that all foreseeable faults have been eliminated before actual execution. Using digital twins or dedicated simulation software, the equipment models, process parameters, and task sequences of the target combination method are imported, and the processing is simulated in chronological order. The movement, signal interactions, and material flow of all devices are monitored. If no interference alarms, scheduling conflicts, or timing competitions occur, the verification is considered successful; otherwise, it fails. After successful simulation verification, control commands are generated based on the target combination method and issued to each device to execute the physical reconfiguration. Simulation verification can identify problems in advance without risk or material waste, improving the success rate and safety of physical reconfiguration.

[0018] It should be noted that if the simulation verification fails, similar to the previous conflict detection, a combination with the second-best performance index needs to be selected as the target combination until the selected target combination passes the simulation verification.

[0019] Interference alarms are shutdown warnings issued when a potential collision is detected. For example, in a production line consisting of automated guided vehicles (AGVs) and lifting platforms, the AGV transports materials to the area below the lifting platform, which then needs to descend to retrieve them. If, during the descent, the AGV fails to completely stop at a safe position due to navigation deviation, and the lifting platform's sensors detect an unidentified object below, an interference alarm will be triggered. Scheduling conflicts refer to the unreasonable scheduling that may occur when multiple tasks are executed in parallel in a combination mode, resulting in tasks vying for the same equipment. Timing competition refers to the situation where, when two devices execute according to a preset time sequence, fluctuations in processing time cause the actual execution order of their access to shared resources (such as the same physical limit switch signal) to differ from the expected order, leading to instruction loss or misjudgment of status.

[0020] In one specific embodiment, after step S5, the following steps are also performed: Receive readiness confirmation signals from each reorganized device to confirm that the new production unit has been successfully established. If the target combination method is a newly generated target combination method, register the newly generated target combination method and the corresponding task type information as a new standard solution in the solution library. Also, store the recorded processing time, conversion time data and the newly generated target combination method in the solution library.

[0021] Specifically, the startup, program loading, and position calibration of physical equipment require a certain amount of time and may fail due to network latency or equipment malfunction. Without a confirmation mechanism, the production process may start prematurely, causing damage to the workpiece. To address the issue of uncertainty about whether equipment has completed reconfiguration after control commands are issued, a readiness confirmation signal is received from each device to ensure all devices are in a working state, thus confirming the successful establishment of the new production unit. While dynamic device combination is flexible, it is computationally expensive. If a new task type occurs repeatedly, re-dynamically combining the devices each time would lead to inefficiency. To solve this problem, after a new production unit is established, if the target combination method is a newly generated target combination method (i.e., the second target combination method mentioned later), the newly generated target combination method and its corresponding task type information are registered as a new standard solution in the solution library. The solution library also associates and stores the actual processing time, conversion time data, and the newly generated target combination method in the production process, facilitating subsequent calls for the same type of task.

[0022] In one specific embodiment, step S3 further includes the following steps: Step S31: Based on the production task information, determine whether there is a standard solution or a combination of multiple standard solutions that can cover all the task types that need to be processed. If so, generate at least one combination of standard solutions, calculate the first index of each combination, calculate the performance index based on the first index, and select the combination with the smallest performance index as the first target combination. If there is only one standard solution that can cover all the task types that need to be processed, select this standard solution as the target combination. Step S32: If not, obtain the standardized capability data and real-time status data of each device, filter out candidate devices that can meet the requirements of each process of the production task based on the standardized capability data, enumerate all feasible device sequences from the candidate device pool based on the process sequence included in the production task, form a set of candidate device combinations, calculate the second index of each candidate device combination based on the target preference data, calculate the performance index based on the second index, and select the device combination with the best performance index as the second target combination method.

[0023] Specifically, a standard solution is an offline optimized combination of equipment and process parameters, and the switching time and processing time between solutions are known. Directly combining standard solutions involves less computation and has a faster response time. To address the problem of how to quickly obtain a feasible production combination when the current production task can be produced using existing standard solutions or combinations of standard solutions, the following steps are taken: First, the task type list in the production task information is read. Then, the standard solution that can produce each task type is queried in the solution library. If each task type has at least one standard solution that can produce it, that is, if there is a standard solution or a combination of standard solutions that can cover all the task types that need to be processed, then based on the process sequence included in the production task, a standard solution is selected from the available standard solutions. The solution library enumerates all feasible solution sequences, forming multiple combinations. For example, if a production task requires processing 50 disc-shaped parts and 30 shaft-shaped parts, the solution library contains solutions 1 and 2 for processing disc-shaped parts, and solutions 3 and 4 for processing shaft-shaped parts. The equipment combination includes four combinations: the combination of solutions 1 and 3, the combination of solutions 1 and 4, the combination of solutions 2 and 3, and the combination of solutions 2 and 4. To select the optimal target combination, the first index of each combination is first calculated. The first index is used as the basic value to measure the efficiency of the combination. After calculating the first index, the final performance index is calculated by combining it with the cycle time penalty. Based on the performance index, the optimal combination is selected as the first target combination.

[0024] It is important to note that if a single standard solution can cover all the task types that need to be processed, then this standard solution should be selected as the target combination method.

[0025] If not found, it indicates that the existing standard solutions do not cover the workpiece types in the production task. To address the issue of constructing a feasible equipment combination when new task types not covered by the solution library arise in the production task, standardized capability data and real-time status data of each device are obtained. Based on the standardized capability data, candidate devices capable of completing each process of the production task are selected. Standardized capability data refers to a pre-set digital description file for each device, which includes numerical parameters and attribute parameters. Numerical parameters include the processing accuracy (values ​​ranging from 0 to 1), unit time cost, and single-piece processing time corresponding to the co-construction type that the device can execute. These numerical parameters are used for subsequent quantitative calculation of performance indicators. Attribute parameters include the communication protocols and data structure specifications supported by the device. These parameters are used for qualitative matching detection when a conflict map is pre-constructed. First, candidate equipment is selected based on the processing accuracy required for the production task. For example, if the production task requires processing 10 shell-type parts, there is no standard solution for this task type. The task sequence is milling, cleaning, and inspection. Milling machines A, B, and E are available for milling, but the processing accuracy of milling machine A does not meet the requirements of the production task. Therefore, milling machines B and E are the preferred options for milling. Cleaning machine C is the preferred option for cleaning, and inspection instrument D is the preferred option for inspection. Both cleaning machine C and inspection instrument D meet the processing accuracy requirements of the production task. This selection is based on the processes included in the production task. The process involves enumerating all feasible equipment sequences from the candidate equipment pool to form a candidate equipment combination set. For example, the generated candidate combination set may include two combinations: a combination of milling machine B, cleaning machine C, and detector D, and a combination of milling machine E, cleaning machine C, and detector D. To select the optimal target combination, a second index is first calculated for each combination. The specific method for calculating the second index will be explained in detail later. Then, based on the second index and the cycle time penalty, the final performance index is calculated. Based on the final performance index, the equipment combination with the optimal performance index is selected as the second target combination.

[0026] In one specific embodiment, calculating a first index for each combination includes the following steps: Step S311: Multiply the number of tasks for each task type by the unit processing time of the standard solution corresponding to that task type, and then sum them to obtain the cumulative processing time; Step S312: The sum of the switching times generated when switching between devices in sequence in the combination mode is taken as the cumulative switching time; Step S313: Add the cumulative processing time and the cumulative switching time to obtain the total production time. Normalize the total production time of all combinations to obtain the normalized total production time index. Use the total production time index as the first index.

[0027] Specifically, the processing quantity of different workpieces varies, and the unit processing time of different standard schemes varies. In order to quantify the actual processing time of all workpieces in the combination method, for each task type, the task quantity is multiplied by the unit processing time of the corresponding standard scheme to obtain the processing time of the corresponding task type. Then, the processing time of each standard scheme is added together to obtain the cumulative processing time. For example, if the production task is to process 50 disc parts and 30 shaft parts, and the selected combination method is scheme 1 and scheme 3, the unit processing time of scheme 1 is 120 seconds, and the unit processing time of scheme 2 is 180 seconds, then the cumulative processing time is 50×120 + 30×180 = 11400 seconds.

[0028] In production with multiple task types, switching between equipment takes up time. If the switching time is not considered, the selected combination may be inefficient in practice due to the frequent switching time. In order to solve the problem of preparation time caused by switching between different standard solutions in the quantitative combination method, the time spent switching equipment each time is accumulated as the non-processing time cost of the corresponding combination method, so as to avoid selecting inefficient solutions with short processing time but frequent switching.

[0029] To address the issue of evaluating the overall efficiency of a combination method, the total production time is obtained by adding the cumulative processing time and the cumulative switching time. To facilitate the subsequent calculation of the final performance index based on the first indicator, the total production time of all combination methods is normalized to obtain the normalized total production time index. The total production time index is used as the first indicator. The first indicator can serve as a clear and easy-to-understand time benchmark. Subsequently, the final performance index of the combination method can be calculated based on the first indicator in combination with other conditions.

[0030] In one specific embodiment, the calculation of a second metric for each candidate device combination based on target preference data includes the following steps: Step S321: Calculate the total production time for each candidate equipment combination; Step S322: Calculate the unit time cost of each piece of equipment and multiply it by the corresponding processing time to obtain the equipment processing cost. Add up the processing costs of each piece of equipment to obtain the total manufacturing cost. Step S323: Calculate the machining accuracy of each device, take the minimum value among the machining accuracies as the machining quality score, and use the value obtained by subtracting the machining quality score as the comprehensive quality score. Step S324: Normalize the total production time, total manufacturing cost, and overall quality score; Step S325: The production time weight, manufacturing cost weight, and processing quality weight in the target preference data are weighted and summed with the normalized total production time, total manufacturing cost, and overall quality score, respectively, and the resulting values ​​are used as the second indicator for each candidate equipment combination.

[0031] Specifically, for newly generated equipment combinations, the corresponding second indicator is calculated as a reference value for evaluating the equipment combination. The second indicator needs to take into account multiple parameters. In order to solve the problem of how to quantify the total processing time required for a equipment combination to complete all processing tasks, the ideal total processing time is obtained by accumulating the working time of each equipment. For a candidate equipment combination, such as D1, D2 and D3, each equipment is responsible for one process. The unit processing time and corresponding processing quantity of each equipment are known. The equipment processing time of each equipment = unit processing time × processing quantity. Then, the equipment processing times of all equipment are added together to obtain the total processing time. The switching time generated when switching equipment in the candidate equipment combination is also calculated. The total production time is obtained by adding all the switching time generated in the candidate equipment combination to the total processing time. Newly generated combinations may be stored as standard solutions in the solution library. Considering only processing time is insufficient, as the operating costs of different equipment vary significantly; for example, high-precision equipment is more expensive. If only production time is considered, a solution with a shorter time frame but higher cost might be selected. To simultaneously consider the total manufacturing cost required for the processing task, the unit time cost of each device is calculated by multiplying it by the corresponding processing time to obtain the device's processing cost. The processing costs of each device are then added together to obtain the total manufacturing cost. To address the issue of quantifying the overall processing quality level of the equipment combination, since the final product quality is limited by the device with the lowest precision in serial processing, the processing precision of each device in the candidate equipment combination is obtained. The minimum value among these precision values ​​is used as the processing quality score. Since the time and cost mentioned earlier are inversely proportional to the final second indicator, to maintain consistency, the value obtained by subtracting the processing quality score from the total quality score is used as the comprehensive quality score. It should be noted that the processing precision is a value between 0 and 1. Since the evaluation parameters for the three dimensions of total production time, total manufacturing cost, and overall quality score are calculated with different dimensions, in order to unify the dimensions, the maximum and minimum values ​​of each evaluation parameter are collected for all candidate equipment combinations. For each evaluation parameter of each candidate equipment combination, a normalized value is calculated. Taking the evaluation parameter of total production time as an example, the calculation method of its corresponding normalized value is: Normalized value (normalized total production time) = (current total production time - minimum total production time among all candidate equipment combinations) / (maximum total production time among all candidate equipment combinations - minimum total production time among all candidate equipment combinations). Since the equipment combinations are generated for the current production task, the production time weight, manufacturing cost weight, and processing quality weight in the current target preference data are weighted and summed with the normalized total production time, total manufacturing cost, and overall quality score, respectively. The value obtained after weighted summation is used as the second indicator for each candidate equipment combination. That is, the second indicator = normalized total production time × production time weight + normalized total manufacturing cost × manufacturing cost weight + normalized overall quality score × processing quality weight.Different production tasks place different degrees of importance on different objectives. If a production task with the same task type but different objective preference data is needed in the future, only the objective preference data needs to be modified to calculate the corresponding second indicator.

[0032] In one specific embodiment, calculating the performance index based on the first index includes the following steps: For each combination method, obtain the baseline beat value of each standard scheme, obtain the maximum and minimum values ​​among all baseline beat values, divide the maximum value by the minimum value to obtain the maximum beat ratio, for combination methods with a maximum beat ratio greater than a preset threshold, add a preset increment value to the corresponding first index, and use the first index after adding the preset increment value as the final performance index of the corresponding combination method, for combination methods with a maximum beat ratio less than or equal to the preset threshold, use the corresponding first threshold as the final performance index.

[0033] Specifically, the first indicator only considers production time but does not take into account the impact of cycle time differences between different standard schemes on actual production. The benchmark cycle time value refers to the single-piece processing time of the slowest equipment under the standard scheme, which limits the overall output speed. If the benchmark cycle time value of one scheme in the combination is large and that of another scheme is small, the short-cycle scheme will frequently wait, resulting in low equipment utilization and wasted energy during standby. Even if the total production time is the same, the combination with balanced cycle time has higher actual output efficiency. Therefore, a cycle time penalty term needs to be added to correct the deviation of the first indicator. To solve the problem that selecting a combination method based solely on the first indicator may result in an inefficient combination with severely mismatched cycle times, leading to frequent waiting in actual production, for each combination method, the benchmark cycle time value of each standard scheme is first obtained. Then, the maximum and minimum values ​​among all benchmark cycle time values ​​are taken, and the maximum value is divided by the minimum value to obtain the maximum cycle time ratio. The beat ratio reflects the speed difference between different standard solutions in a given combination. If the maximum beat ratio of a combination is greater than a preset threshold, it indicates that the beat ratios of the standard solutions in the combination are not well matched, leading to frequent waiting for one solution and low equipment utilization. The preset threshold can be set by the staff according to the task situation, for example, it can be set to 2. For combinations with mismatched beat ratios, the corresponding first indicator is increased by a preset increment value. Assuming the preset value is 0.5 and the first indicator for beat mismatch is 0.3, then the final performance indicator of the combination with mismatched beat ratios is 0.3 + 0.5 = 0.8. Increasing the preset value improves the performance indicator of the corresponding combination with mismatched beat ratios. For combinations with a maximum beat ratio less than or equal to the preset threshold, since their beat ratios are relatively matched, there is no need to increase their performance indicator; the corresponding first indicator is used as the final performance indicator of the corresponding combination. In this way, the performance indicator of combinations with mismatched beat ratios is increased, while the performance indicator of combinations with matched beat ratios is not increased. This makes it easier to select the better combination based on the performance indicator.

[0034] Furthermore, performance metrics are calculated based on the second metric, including: For each candidate equipment combination, obtain the single-piece processing time of each equipment, obtain the maximum and minimum values ​​among all single-piece processing times, divide the maximum value by the minimum value to obtain the maximum cycle time ratio, and for combinations with a maximum cycle time ratio greater than a preset threshold, add a preset value to the corresponding second indicator, and use the obtained value as the final performance indicator of the corresponding candidate equipment combination.

[0035] Specifically, the method for calculating performance indicators based on the second indicator is the same as the method for calculating performance indicators based on the first indicator, but the processing time of each piece of equipment in the equipment combination is selected when obtaining the data.

[0036] In one specific embodiment, the pre-constructed conflict map includes the following steps: Step S411: Identify and list the possible conflict types between multiple devices; Step S412: Generate corresponding conflict information for each conflict type. The conflict information includes the first device identifier, the second device identifier, the conflict type code, and the severity flag, which includes tolerable and intolerable categories. Step S413: Store all conflict information as a conflict list. The first row and first column of the conflict list correspond to all device identifiers, and each cell at the intersection of the row and column records the conflict type code and severity mark of the two corresponding devices.

[0037] Specifically, to address the need to identify which types of conflicts need to be detected during rapid pre-screening, the potential conflict types between multiple devices are first identified and listed. These conflict types include spatial overlap conflicts, communication protocol incompatibility, and auxiliary resource contention conflicts. To address the issue of how to structurally record conflicts for easy querying and severity assessment, corresponding conflict information is generated for each conflict type. This conflict information includes the identifiers of the first and second devices involved in the conflict, the conflict type code, and a severity flag. The severity flag is categorized as tolerable or intolerable, allowing for direct retrieval during subsequent rapid detection. Tolerable conflicts refer to those that can be resolved through cost-effective methods. Conflicts that cannot be resolved by other means, such as incompatible communication protocols between the robot and the detector, can be resolved by adding a protocol conversion gateway. Intolerable conflicts refer to unsolvable conflicts between two devices, such as mismatched material interface sizes or inability to connect the output port of one device to the input port of another, which cannot be resolved by installing adapters. To ensure efficient querying of conflict information between devices, all conflict information is stored as a conflict list, equivalent to a two-dimensional matrix. The first row and first column of the conflict list correspond to all device identifiers, and each cell at the intersection of the row and column records the conflict type code and severity flag for the corresponding two devices. This method constructs a conflict graph, facilitating subsequent querying, maintenance, and expansion.

[0038] In one specific embodiment, the detection of target combination methods based on the conflict map includes the following steps: Step S421: Combine the device identifiers of all adjacent devices included in the target combination method to form all device pairs that need to be detected; Step S422: For each device pair, query the conflict map. If the corresponding conflict information is found, obtain the corresponding severity label. Step S423: If any device pair has a conflict marked as intolerable in severity, the target combination method is determined to have failed the test. If none of the device pairs have an intolerable conflict, the target combination method is determined to have passed the test.

[0039] Specifically, since only two adjacent devices need to be connected, in order to ensure that no potential conflict between any pair of devices is missed, all device pairs that need to be detected are formed by using the device identifiers of all adjacent devices included in the target combination method. For each device pair, the conflict map is queried to see if there is corresponding conflict information. If so, the corresponding severity label is obtained. If any device pair has a conflict with a severity label of intolerable, the target combination method is determined to have failed the test. If no device pair has an intolerable conflict, the target combination method is determined to have passed the test.

[0040] Furthermore, when generating device pairs, the following is also performed: For the first target combination method, if the first target combination method contains more than two standard schemes, only the equipment pairs that need to be connected in the adjacent standard schemes of the first target combination method will be generated.

[0041] To more quickly detect target combinations based on the conflict map, when generating device pairs, for the first target combination, if the first target combination contains more than two standard schemes, since the conflict between devices has already been considered when generating the standard schemes, it is not necessary to perform conflict detection on the two adjacent devices within the standard schemes. It is only necessary to perform conflict detection on the device pairs that need to be connected in the adjacent standard schemes. Since the second target combination is newly generated, it is necessary to perform conflict detection on each pair of adjacent device pairs in the second target combination.

[0042] The above describes the dynamic combination method of flexible intelligent manufacturing units based on modular design in the embodiments of this application. The following describes the dynamic combination system of flexible intelligent manufacturing units based on modular design in the embodiments of this application. Please refer to [link / reference]. Figure 2 One embodiment of the flexible intelligent manufacturing unit dynamic combination system based on modular design in this application includes: The pre-configuration module is used to pre-configure multiple standard solutions. Each standard solution corresponds to a task type. The standard solution contains fixed equipment combinations, process parameters, processing time, conversion time data, and baseline cycle time values. The task acquisition module is used to acquire production task information, which includes the type of task to be processed, the number of tasks of each type, and target preference data. The target preference data includes production time weight, manufacturing cost weight, and processing quality weight. The combination optimization module is used to generate at least one combination method based on production task information, standard schemes and standardized capability data of each equipment, calculate the performance index of each combination method, select the combination method with the best performance index as the target combination method, and select this combination method as the target combination method when only one combination method is generated. The conflict detection module is used to detect target combination methods based on a pre-built conflict map. If there is a conflict in the target combination method, the combination method with the second best performance index is selected as the target combination method until the selected target combination method has no conflict. The simulation deployment module is used to simulate the operation of the selected target combination in a virtual environment to verify its feasibility. After the simulation verification is successful, control commands are generated according to the target combination and sent to each device to execute the physical layer reorganization.

[0043] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A dynamic combination method of flexible intelligent manufacturing units based on modular design, characterized in that, The method includes: Step S1: Pre-configure multiple standard schemes. Each standard scheme corresponds to a task type. The standard scheme contains fixed equipment combinations, process parameters, processing time, conversion time data, and baseline cycle time values. Step S2: Obtain production task information. Production task information includes the type of task to be processed, the number of tasks of each type, and target preference data. Target preference data includes production time weight, manufacturing cost weight, and processing quality weight. Step S3: Based on the production task information, standard scheme and standardized capability data of each equipment, generate at least one combination method, calculate the performance index of each combination method, select the combination method with the best performance index as the target combination method, and select this combination method as the target combination method if only one combination method is generated. Step S4: Detect the target combination method based on the pre-constructed conflict map. If there is a conflict in the target combination method, select the combination method with the second best performance index as the target combination method until there is no conflict in the selected target combination method. Step S5: Simulate the operation process of the selected target combination method in a virtual environment to verify its feasibility. After the simulation verification is passed, generate control commands according to the target combination method and send control commands to each device to execute the physical layer reorganization.

2. The method of claim 1, wherein, Following step S5, the following is also included: Receive readiness confirmation signals from each reorganized device to confirm that the new production unit has been successfully established. If the target combination method is a newly generated target combination method, register the newly generated target combination method and the corresponding task type information as a new standard solution in the solution library. Also, store the recorded processing time, conversion time data and the newly generated target combination method in the solution library.

3. The method of claim 1, wherein, Step S3 includes: Step S31: Based on the production task information, determine whether there is a standard solution or a combination of multiple standard solutions that can cover all the task types that need to be processed. If so, generate at least one combination of standard solutions, calculate the first index of each combination, calculate the performance index based on the first index, and select the combination with the smallest performance index as the first target combination. If there is only one standard solution that can cover all the task types that need to be processed, select this standard solution as the target combination. Step S32: If not, obtain the standardized capability data and real-time status data of each device, filter out candidate devices that can meet the requirements of each process of the production task based on the standardized capability data, enumerate all feasible device sequences from the candidate device pool based on the process sequence included in the production task, form a set of candidate device combinations, calculate the second index of each candidate device combination based on the target preference data, calculate the performance index based on the second index, and select the device combination with the best performance index as the second target combination method.

4. The method of claim 3, wherein, Calculate the first metric for each combination, including: Step S311: Multiply the number of tasks for each task type by the unit processing time of the standard solution corresponding to that task type, and then sum them to obtain the cumulative processing time; Step S312: The sum of the switching times generated when switching between devices in sequence in the combination mode is taken as the cumulative switching time; Step S313: Add the cumulative processing time and the cumulative switching time to obtain the total production time. Normalize the total production time of all combinations to obtain the normalized total production time index. Use the total production time index as the first index.

5. The method of claim 3, wherein, A second metric is calculated for each candidate device combination based on target preference data, including: Step S321: Calculate the total production time for each candidate equipment combination; Step S322: Calculate the unit time cost of each piece of equipment and multiply it by the corresponding processing time to obtain the equipment processing cost. Add up the processing costs of each piece of equipment to obtain the total manufacturing cost. Step S323: Calculate the machining accuracy of each device, take the minimum value among the machining accuracies as the machining quality score, and use the value obtained by subtracting the machining quality score as the comprehensive quality score. Step S324: Normalize the total production time, total manufacturing cost, and overall quality score; Step S325: The production time weight, manufacturing cost weight, and processing quality weight in the target preference data are weighted and summed with the normalized total production time, total manufacturing cost, and overall quality score, respectively, and the resulting values ​​are used as the second indicator for each candidate equipment combination.

6. The method of claim 3, wherein, Performance metrics are calculated based on the first metric, including: For each combination method, obtain the baseline beat value of each standard scheme, obtain the maximum and minimum values ​​among all baseline beat values, divide the maximum value by the minimum value to obtain the maximum beat ratio, for combination methods with a maximum beat ratio greater than a preset threshold, add a preset increment value to the corresponding first index, and use the first index after adding the preset increment value as the final performance index of the corresponding combination method, for combination methods with a maximum beat ratio less than or equal to the preset threshold, use the corresponding first threshold as the final performance index.

7. The method of claim 1, wherein, The pre-constructed conflict map includes: Step S411: Identify and list the possible conflict types between multiple devices; Step S412: Generate corresponding conflict information for each conflict type. The conflict information includes the first device identifier, the second device identifier, the conflict type code, and the severity flag, which includes tolerable and intolerable categories. Step S413: Store all conflict information as a conflict list. The first row and first column of the conflict list correspond to all device identifiers, and each cell at the intersection of the row and column records the conflict type code and severity mark of the two corresponding devices.

8. The method of claim 1, wherein, Detection of target combination methods based on conflict maps, including: Step S421: Combine the device identifiers of all adjacent devices included in the target combination method to form all device pairs that need to be detected; Step S422: For each device pair, query the conflict map. If the corresponding conflict information is found, obtain the corresponding severity label. Step S423: If any device pair has a conflict marked as intolerable in severity, the target combination method is determined to have failed the test. If none of the device pairs have an intolerable conflict, the target combination method is determined to have passed the test.

9. A flexible smart manufacturing cell dynamic combination system based on modular design for implementing the flexible smart manufacturing cell dynamic combination method based on modular design according to any one of claims 1-8, characterized in that, The system includes: The pre-configuration module is used to pre-configure multiple standard solutions. Each standard solution corresponds to a task type. The standard solution contains fixed equipment combinations, process parameters, processing time, conversion time data, and baseline cycle time values. The task acquisition module is used to acquire production task information, which includes the type of task to be processed, the number of tasks of each type, and target preference data. The target preference data includes production time weight, manufacturing cost weight, and processing quality weight. The combination optimization module is used to generate at least one combination method based on production task information, standard schemes and standardized capability data of each equipment, calculate the performance index of each combination method, select the combination method with the best performance index as the target combination method, and select this combination method as the target combination method when only one combination method is generated. The conflict detection module is used to detect target combination methods based on a pre-built conflict map. If there is a conflict in the target combination method, the combination method with the second best performance index is selected as the target combination method until the selected target combination method has no conflict. The simulation deployment module is used to simulate the operation of the selected target combination in a virtual environment to verify its feasibility. After the simulation verification is successful, control commands are generated according to the target combination and sent to each device to execute the physical layer reorganization.

Images